Thanks for the complement (inhibitor)

Charles Parker

Developing an effective inhibitor of the alternative pathway of complement (APC) with potential for safe clinical use is a daunting undertaking, but Fridkis-Hareli and colleagues were up for the challenge.1

The straightforward strategy of generating an antibody that inhibits complement C3, the protein that serves as the nidus for formation of the APC C3 convertase (see figure) is doomed from the beginning because C3 is present in high concentration in the plasma and has a rapid turnover rate that is enhanced by inflammation. Further, continuous depletion of C3 would be associated with a high degree of morbidity and mortality as the protein plays a central role in protection of humans against bacterial infection and immune complex disease.2 To circumvent these intimidating obstacles, Fridkis-Hareli and colleagues used a cleverly designed chimeric, recombinant protein (TT30) that targets C3 selectively at the site of activation of the APC (see figure).

TT30 is composed of 2 functional domains derived from distinct complement proteins (see figure).3 One component of the chimeric protein functions as the recognition subunit, and the other serves as the inhibitory subunit. The recognition subunit is derived from complement receptor type 2 (CR2). CR2 (CD21) is a cellular receptor for degradation products of C3 (iC3b and C3dg) that are generated at sites of complement activation (see figure). The inhibitory component of TT30 is a truncated version of complement factor H, the primary plasma inhibitor of the APC. Factor H binds to activated C3 (C3b) and both prevents binding of complement factor B and decay-accelerates the catalytic subunit (activated factor B) from the APC C3 convertase.4 In addition, factor H acts as a cofactor for the degradation of C3b to inactive iC3b by complement factor I (see figure). The idea behind design of the chimeric protein was that the CR2 domain would localize binding to sites of APC activation, positioning the factor H component to bind to nascent C3b, thereby preventing APC amplification. A unique feature of C3 is that when activated, an internal thioester bond is exposed that allows the molecule to bind covalently to membrane glycoproteins through ester bond formation (see figure).5 In this way, activated C3b becomes fixed to the surface of the structure (eg, the red cell membrane) on which complement has been activated. The very short half-life (milliseconds) of the exposed thioester bond restricts the diffusion capacity of activated C3, resulting in clustering of C3b around a C3 convertase. Such a process creates a favorable microenvironment for an inhibitor like TT30, as the molecule would be positioned to reach multiple molecules of C3b within the cluster.

Using a broad, rigorous, experimental design, Fridkis-Hareli and colleagues investigated the capacity of TT30 to inhibit specifically the APC.1 Ex vivo studies demonstrated that TT30 was a potent inhibitor of the APC with activity observed in the nanamolar concentration range. Specificity was shown in ex vivo experiments in which the inhibitory capacity of TT30 for the APC was found to be approximately 100-fold greater than for the antibody-dependent classical pathway of complement (CPC).5 Another set of ex vivo experiments produced the remarkable finding that the APC inhibitory activity of TT30 was 150-fold greater than that of factor H, the primary plasma regulator of the APC. This observation supports Fridkis-Hareli and colleagues' hypothesis that the potency of factor H would be enhanced by targeting the protein to the site of APC activation through the CR2 binding domain.

When incubated in normal human serum, rabbit erythrocytes undergo explosive spontaneous hemolysis because the sialic acid recognition site for factor H binding to C3b is deficient (see figure). Fridkis-Hareli et al used this system to demonstrate the capacity of TT30 to block APC-mediated hemolysis. They also demonstrated by flow cytometry that TT30 remained bound to rabbit erythrocytes for more than 24 hours. This observation has important clinical implication as binding of TT30 to covalently bound C3 fragments on erythrocytes could be recognized as a foreign complex (see figure), thereby generating an antibody response that would lead to immune-mediated red cell destruction.

Illustration of covalent binding of activated C3 (C3b) to glycophorin A (GPA) on the erythrocyte membrane surface. The bound C3b serves as the nidus for formation of the APC C3 convertase (C3b, activated factor B [Bb], and factor P) that enzymatically activates many molecules of C3 to C3b that then bind covalently via an exposed thioester bond to carbohydrate residues on GPA. Supported by interaction with sialic acid residues on GPA, the plasma protein, factor H, binds to C3b and serves as a cofactor for degradation of C3b to iC3b by the plasma protein factor I. Complement receptor I (CR1) also binds to C3b and to iC3b and serves as a cofactor for degradation of C3b to iC3b and then C3dg by factor I. TT30 binds to both iC3b and C3dg through its CR2 domain (red circles). This binding positions the factor H–derived inhibitory component of TT30 (blue circles) to interact with nascent C3b molecules generated by an active C3 convertase.

The pharmacodynamic and pharmacokinetic properties of TT30 were studied in cynomolgus monkeys. Bioavailability was demonstrated for both intravenous and subcutaneously injected TT30 with a 3-fold longer duration of complete APC inhibition when injected subcutaneously (24 hours) compared with intravenously (8 hours). Moderate inhibition (60%) of the CPC of relatively short duration (4 hours) was observed for intravenously infused TT30. Immunohistochemical staining of cryosections from healthy and diseased (asthmatically inflamed) lungs demonstrated localization of TT30 to sites of complement activation, supporting the hypothesis that TT30 would be active in controlling APC activation in solid organs.

These studies of Fridkis-Hareli and colleagues have demonstrated the development of a selective inhibitor of the APC with potential for clinical use. But in what pathologic conditions might TT30 be beneficial? An obvious candidate is paroxysmal nocturnal hemoglobinuria (PNH), a disease in which hemolysis is the result of uncontrolled activation of the APC in combination with lack of inhibition of formation of the cytolytic complement membrane attack complex because of deficiency of decay-accelerating factor (CD55) and membrane inhibitor of reactive lysis (CD59), respectively.6 But is not a safe, effective inhibitor of complement-mediated hemolysis of PNH erythrocytes available in the form of the humanized monoclonal anti-complement C5 antibody eculizumab (Soliris; Alexion Pharmaceuticals Inc)?7 Yes, but recent studies have suggested that some PNH patients respond suboptimally to eculizumab because APC formation is not affected by blocking C5, resulting in opsonization of PNH erythrocytes by activation and degradation products of C3 and subsequent extravascular hemolysis mediated by complement receptors on reticuloendothelial cells.8 Hypothetically, TT30 would prevent both intravascular and extravascular hemolysis of PNH erythrocytes by inhibiting formation of the APC C3 convertase, thereby eliminating downstream formation of the cytolytic membrane attack complex. In addition, atypical hemolytic uremic syndrome (a disease that also responds to eculizumab) would be a potential target for TT30 as aberrant regulation of the APC because of inherited mutations of components of the APC underlies the pathobiology of this disease.9,10 Conversely, chronic cold agglutinin disease would not be expected to respond to TT30 as activation of the CPC by IgM antibody accounts for the complement-mediated hemolysis that characterizes this disease.

Based on the rigorous preclinical studies of Fridkis-Hareli et al,1 TT30 warrants further development. Determining whether the efficacy of TT30 will be limited by antibody formation against the recombinant protein itself or against neoepitopes created by binding of TT30 to covalently bound C3 degradation product will require further investigation. But to paraphrase Mark Twain, a complaint should always be preceded by a compl(e)ment.


  • Conflict-of-interest disclosure: The author declares no competing financial interests. ■